Does this make a difference in GB, on average.?.
I was talking of the MT score, to get back on topic Strix Point@30W score about 13600 pts in the less than ideal GB 6.
I edited my post aboveDoes this make a difference in GB, on average.?.
Ha missed that these were different SoC. Too many variables here.Congrats! You're comparing a machine with 5 or 6 P-cores and 6 E-cores to one with a minimum of ten P-cores and 4 E-cores and declaring there's no way the latter has higher MT perf.
Completely unrelated, to anyone following along who wants to improve their reading comprehension, there are a lot of great resources for that these days.
Unless I missed something M4 is still 4x 128 according to Anandtech's own article on X925.Apple and Cortex-x925 with its six 128bit NEON pipes
That's great advice for those that don't want to read that stuff.To guys dirtying the thread with this irrelevant apple yada yada thing, welcome to my ignore list! Now Zen 5 thread is clean
yeah, thing is Apple is just a niche, not a competitor for the products I buy. Apple is not an option nor alternative to me.That's great advice for those that don't want to read that stuff.
These products do not exist in a vacuum, so you have to accept there will be compare and contrast. But as you have demonstrated, the ignore feature is a great tool for curating your experience here.
This is best discussed in the architectural thread I think.
spectacular branch prediction improvements, bigger chips like Turin and zen6 will reap the benefits
That’s not the point tho, we are talking about Strix Point which is mobile and Apple competes there. If it was desktop yes Apple is irrelevant there. It’s just a chip and architecture comparison.yeah, thing is Apple is just a niche, not a competitor for the products I buy. Apple is not an option nor alternative to me.
Yes we have gone a bit far, I admit and apologize. But there still was some information to pick and little trolling. Wearing blinder is a poor way to get fair information.That’s not the point tho, we are talking about Strix Point which is mobile and Apple competes there. If it was desktop yes Apple is irrelevant there. It’s just a chip and architecture comparison.
This is also true of AMD themselves, they compare with Apple on mobile and ignore them on desktop. All this nonsense wouldn’t have come about if the discussion was only about the chips and how different architectures compared but instead we trailed off.
Man, the cross-CCX penalty is 3 times higher than that of Renoir. It's atrocious.
Think about it. AMD is becoming a software company. What if they get AI to write their compilers and some other widely used open source libraries to make the maximum use of their architectures?
It doesn't.How much does this matter outside of applications or the OS not being aware of the P/E cores?
No, 4C is becoming the standard from either vendor.Ideally AMD will jump up to 6 P-cores in future APUs so that there's less need to utilize their c-cores when the regular cores become saturated.
I distinctly remember a time when I was 7 years old and dropped such a "what if" on one of my cousins, who promptly responded with "What if monkeys flew out of my butt?"
oh yes, its the point to me. Apple and its closed ecosystem isnt for me, at all. So, it cant care me less, as doesnt care to most people btw.That’s not the point tho, we are talking about Strix Point which is mobile and Apple competes there. If it was desktop yes Apple is irrelevant there. It’s just a chip and architecture comparison.
This is also true of AMD themselves, they compare with Apple on mobile and ignore them on desktop. All this nonsense wouldn’t have come about if the discussion was only about the chips and how different architectures compared but instead we trailed off.
fair enoughoh yes, its the point to me. Apple and its closed ecosystem isnt for me, at all. So, it cant care me less, as doesnt care to most people btw.
Depends entirely what you mean by "modern".Who enabled the modern ARM push? Single handedly Apple
AVX512 is a failure because its support has been a fragmented mess from the beginning.AVX-512 is also a failure because it's based on the Old Intel ideology where they can make general purpose CPUs fast enough to stem the tide of GPUs forever
Literally the entire point of it is to fix AVX support in heterogeneous CPU client SoCs.I wouldn't be surprised if the E core team stays away from AVX10.1/10.2 as long as possible.
Pretty much this. Idk why are they so obsessed with "muhhh Apple did it better"oh yes, its the point to me. Apple and its closed ecosystem isnt for me, at all. So, it cant care me less, as doesnt care to most people btw.
Pretty much this. Idk why are they so obsessed with "muhhh Apple did it better"
> cant even run the majority of the meaningful x86 applications other than the ones that have been ported to apple silicon
> completely different targets of applications and instructions
> No meaningful gains after M1 (except the fancy marketing graph that they put up)
I think M4 triumphalism is a bit early. Of course it'll be even faster when it shows up in MacBooks (presumably this fall). But that hasn't happened yet. I can understand why they're not swayed at all by a processor as of yet only in iPads.
And if my math is correct AMD is actually slightly closer to Apple now than they were in the fall of 2020.
That's true. People keep obsessing over IPC when it is only one component (albeit a very nice component) of performance.Okay. Point stands even with M3. That's +28% since M1, plus small-moderate iso clock gains - hardly "no meaningful gains"